Professional Documents
Culture Documents
cytokine, in peripheral blood mononuclear cells (PBMC) but not to the ability to block
IL-8 secretion in TNF-α-stimulated HT-29 cells, reinforcing the hypothesis of a complex
anti-inflammatory pathway driven by F. prausnitzii. Altogether, our results suggest that
some F. prausnitzii strains could represent good candidates as next-generation probiotic.
Keywords: probiotic, commensal, Faecalibacterium, molecular and metabolic characterization,
immune-modulatory properties
TABLE 1 | Studied cohort of healthy humans’ volunteers and new F. prausnitzii strain identified.
Subject Sex Age (years) Fecal SCFA (mM) CFU/g % EOS Identified F. prausnitzii strains Cultivability of the strain
All the isolates were obtained from human fecal samples of healthy volunteers consuming omnivorous diets. F, female; M, male; nd, not determined, X, no identified F. prausnitzii strain.
supplemented with rumen fluid 20%. After 4 days of incubation (DMEM) (Sigma-Aldrich) supplemented with 10% (w/v) heat-
at 37◦ C, single colonies were obtained on plates and 96 varied inactivated fetal bovine serum (FBS) (GibcoBRL, Eragny, France)
colonies were selected and isolated in duplicate on YHBHI and with penicillin G/ streptomycin (5,000 IU/mL, 5,000 µg/mL)
supplemented with rumen fluid 20% agar plate. A group of (Sigma-Aldrich). Cultures were incubated in 25 cm2 tissue
plates was placed brought out of the anaerobic chamber for culture flasks (Nunc, Roskilde, Denmark) at 37◦ C in a 5% (v/v)
1 h to eliminate EOS strains and after a long period of CO2 atmosphere until confluence.
incubation (usually between 48 h and 4 days), we performed a
negative screening. The EOS colonies were further re-isolated
and a specific F. prausnitzii PCR (primers Fprau07/Fprau02) was
16S rRNA Gene Analysis
done to identify strains of this specie (Table 2). Finally, a 16S DNA was extracted from isolated colonies of the different F.
rRNA gene sequencing was performed after complete 16S rRNA prausnitzii strains by alkaline lysis in 50 µL of NaOH 0.5 M
amplification using primer FP1 to FP5 (Table 2; MWG France). during 30 min and 50 µL of Tris 1M pH7 and 100 µL H2 O were
The viable isolates were stocked at −80◦ C with 16% of glycerol. added. 16S rRNA sequences were amplified using FP1 and FP2
primers (Table 2) and PCR products purified with the Wizard
SV Gel. PCR Clean-Up system (Promega) was used to obtain
Bacterial Strains, Cell Culture, and Growth bidirectional partial 16S rRNA gene sequences by using primers
Conditions FP1, FP2, FP3, FP4, and FP5 (Table 2). All DNA sequences were
The reference strains A2-165 (DSM 17677; Duncan et al., 2002), confirmed by sequencing (Eurofins MWG Operon, Ebersberg,
L2/6 (Barcenilla et al., 2000) and M21/2 (Louis et al., 2004) and Germany). Sequences for the novel isolates were deposited in
the F. prausnitzii isolated strains (Table 1) were grown at 37◦ C in the NCBI database under the accession numbers MF185398 to
YBHI medium supplemented with cellobiose (1 mg/ml; Sigma), MF186168.
maltose (1 mg/ml; Sigma), and cysteine (0.5 mg/ml; Sigma) in Phylogenetic analysis based on 16S rRNA were performed
an anaerobic chamber filled with N2 = 90%, CO2 = 5% and using the multiple sequence alignment—CLUSTALW
H2 = 5%. (Thompson et al., 1994) integrated in MEGA6 software (Tamura
HT-29 (ATCC HTB-38) (LGC-Standars) cell line was grown et al., 2013). After that, the most appropriate evolutionary
in Dulbecco’s Modified Eagle’s minimal essential medium model was defined and the evolutionary history was inferred
Primer Oligonucleotide sequence (5′ –3′ ) PCR product size (bp) Use References
using the Maximum likelihood (ML) criterion, based on the conditions recommended by the supplier (Biomerieux, France).
Kimura 2-parameter model (Kimura, 1980), with 1,000 bootstrap The results were recorded after 48 h of incubation.
replicates. A discrete Gamma distribution was used to model
evolutionary rate differences among sites [five categories (+G, Anti-bacterial Assays
parameter = 0.1846)]. The rate variation model allowed for The anti-bacterial effect of F. prausnitzii supernatants were
some sites to be evolutionarily invariable ([+I], 64.70% sites). investigated in vitro using the bacteriocin activity assay as
Initial tree(s) for the heuristic search were obtained by applying previously described (Ramirez-Farias et al., 2009). This anti-
the Neighbor-Joining method to a matrix of pairwise distances bacterial effect was tested on six different bacterial species:
estimated using the Maximum Composite Likelihood (MCL) three aerobic bacteria (E. coli Nissle 1917, E. coli DH10B,
approach and all positions containing gaps and missing data were and Listeria monocytogenes 11765), one facultative anaerobic
eliminated. The tree with the highest log likelihood (−3073.67) is bacterium (Lactococcus subsp cremoris MG1363), and two
shown (Figure 3). The tree is drawn to scale, with branch lengths obligate anaerobic bacteria (Clostridium perfringens ATCC13124
measured in the number of substitutions per site. The analysis and Bifidobacterium infantis DSM20088/ATCC15697). YBHI
involved 36 nucleotide sequences. There were a total of 1090 liquid medium alone was used as negative control.
positions in the final dataset. In this analysis, sequences used
by Lopez-Siles et al. (Duncan et al., 2002; Ramirez-Farias et al., Metabolic Activities
2009; Lopez-Siles et al., 2012) were included with the objective To determine the metabolic activities of the cultivable strains,
of compare the new strains to the two phylogroups proposed by API-20A galleries and the gelatin degradation test of API-20E
that study. Eubacterium desmolans was used to root the tree. galleries were used according to manufacturer’s instructions.
For detection of DNase and hemolytic activity, the strains were
grown ON and then plated into Methyl green-DNA agar plates
Plasmid Presence
(Difco) or blood agar plates (Biomérieux) respectively. The
The presence of plasmids in the isolated strains were determined
results were recorded after 48 h of incubation. The capacity
following Wizard
R
Plus SV Minipreps DNA Purification System
to grow in presence of mucin was assayed using a defined
(Promega) with modifications to adapt it for use with Gram
medium (KH2 PO4 : 5.236 g/L, (NH4 )2 SO4 : 4 g/L, NaCl: 4 g/L,
positive bacteria. Briefly, an extra lysis step was performed after
CaCl2 : 30 mg/L, MgCl2 : 300 mg/L, MnCl2 : 30 mg/L, FeCl2 :
centrifugation of liquid overnight (ON) cultures by incubation
8 mg/L, Vitamin B12: 5 mg/L, Vitamin B1: 1 mg/L, Biotin: 1 mg/L,
for 1 h at 37◦ C with lysozyme (Sigma; 10 mg/ml) in the cell
PABA: 1 mg/L, Folic acid: 1 mg/L, Vitamin K: 2 mg/L, cystein
resuspension solution.
0.5 mg/mL) supplemented with 1.5% mucin (Type II, Sigma-
Aldrich).
Scanning Electron Microscopy
Scanning electron microscopy analyses were performed on the Short Chain Fatty Acid (SCFA) Analysis
MIMA2 platform (INRA, France) with pure pellet of bacterial Supernatant concentrations of propionate, acetate, and butyrate
culture suspended and fixed in 200 µL of glutaraldehyde and 3% were analyzed using gas liquid chromatography (Nelson 1020,
ruthenium red during 2 h in an anaerobic chamber and stored at Perkin-Elmer, St Quentin en Yvelines, France) as previously
4◦ C. Scanning electron microscopy was performed as previously described (Lan et al., 2008). Overnight culture (20 h) of F.
reported (Joly et al., 2010). prausnitzii strains were used and culture media as negative
control; each measurement for performed at least in triplicate
Determination of Antibiotics Resistance except for fecal samples. SCFA concentrations are expressed
The minimum inhibitory concentrations (MIC) for 13 antibiotics in mM.
(including tetracycline, kanamycin, chloranphenicol, linezolid,
nupri/dalfopri, trimethoprim, gentamicin, erythromycin, Dosage of D- and L-Lactate
cefpirome, clindamycin, streptomycin, vanomycin, and D- and L-lactate was measured in supernatant of bacterial
ampicillin) were determined on Wilkins-Chalgren agar (Difco) cultures. This supernatant was precipitated with trichloroacetic
according to the E-test procedure, in accordance with the acid (10%) and centrifuged at 4,500 g for 20 min at 4◦ C. Lactate
was then measured in the supernatants with the Biosentec Experiments on Peripheral Blood
D/L lactic acid enzymatic kits according to the manufacturer Mononuclear Cells (PBMCs)
instructions (Biosentec, Toulouse, France). Overnight culture The protocol used in this study was adapted from Kechaou
(20 h) of F. prausnitzii strains were used and culture media as et al. (2012). Commercial PBMCs (StemCell Technologies,
negative control; each measurement was performed at least in France) from five healthy donors were used in this assay.
triplicate. Donors presented the following characteristics: men, age under
65, body mass index <30, non-smoking, no drugs with anti-
inflammatory known effects taken during the 15 days prior
Adhesion Assays to sampling, and tested negative for HIV, hepatitis A and B
Monolayers of HT-29 cells were seeded in 24-well tissue culture viruses. After reception, cells were stored in liquid nitrogen
plates (Nunc) with 1.83 × 105 HT-29 cells/well and cultivated until use. To prepare PBMCs for co-culture experiments with
until confluence, culture medium was changed daily. Monolayers bacteria, the vial were thawed at 37◦ C in a water bath and
were then infected in 1 ml of the cell culture medium without then transferred into a medium containing RPMI-1640 medium
antibiotics and with heat-inactivated FBS at a multiplicity of supplemented with 10% heat-inactivated FCS, 1% L-glutamine
infection (MOI) of 100 bacteria per epithelial cell. After, 3 h and 0.1% Penicillin/Streptavidin (medium components were
of incubation at 37◦ C in anaerobic conditions (as describe bought from Lonza, Switzerland). DNase (100 µg/mL, Roche
above), monolayers were washed three times in phosphate- Applied Science, France) was added to this mix to avoid cell
buffered saline (PBS; pH 7.2). The epithelial cells were then lysed clumping. Cells were then centrifuged at 200 g for 15 min,
with 1% Triton X-100 (Sigma Chemical Company, St Louis, counted using trypan blue and spread on 24-well plates at 1 × 106
Mo.) in water. Samples were plated onto YHBHI supplemented cells/well. Supernatants were added in triplicates (three wells per
agar plates to determine the number of CFU corresponding donor) at 10% in a total volume of 1 ml. Plates were incubated for
to the total number of cell-associated bacteria. Adhesion to 24 h at 37◦ C with 10% CO2 . Culture supernatant were collected,
mucin has been performed as previously described by Radziwill- mixed with an antiprotease cocktail according to manufacturer’s
Bienkowska et al. (2014, 2016) Briefly, after an overnight coating instructions (Complete EDTA-Free protease inhibitor, Roche
of 96 plates (Nunc) with a solution of 10 mg/ml of mucin [Type Applied Bioscience) and stored at −80◦ C until further analysis
III mucin from porcine stomach (lyophilized powder, Sigma- of IL-10 concentrations by ELISA (Mabtech, Sweden).
Aldrich)] a bacterial suspension (OD600nm = 1) in PBS of each
strain was incubated 3-h at 37◦ C in the anaerobic chamber.
Bound cells were stained with crystal violet. Stained bacteria were
Statistical Analysis
suspended in 96% ethanol and optical density was determined GraphPad software (GraphPad Sofware, La Jolla, CA, USA) was
at 583 nm. All the experiments were performed in triplicate. used for statistical analysis. Results are presented as bar graphs
The adhesion values have been normalized using Lactobacillus ±SEM. Comparisons were realized with the non-parametric
rhamnosus GG (LGG) a positive control know by their good Kruskal–Wallis test followed by a Dunn’s Multiple Comparison
adhesion properties to mucin (Martin et al., 2015). Results are test. Correlation test were performed using spearman test. A p <
presented by the mean and the standard deviation. 0.05 was considered significant.
FIGURE 2 | Growth profile of F. prausnitzii strains. (A) OD600 nm determination after 20 h growth in YBHI supplemented medium and (B) determination of viable
bacteria: the CFU/mL numeration in the same cultures. Each measurement have been done at least in triplicate.
FIGURE 3 | Phylogenetic tree of F. prausnitzii strains based on 16S rRNA gene sequences. The tree was constructed with the MEGA6 software package using the
Maximum Likelihood method. The bootstrap values above 70% are shown next to the branches. The F. prausnitzii isolates incorporated in this study have circles
besides. The black circles represent the cultured strains and white circles represent uncultured isolates. Colors (purple, blue, and green) and letters (A, B, and C)
indicate the tree groups with high bootstrap values, formed by our cultured strains.
stable against most plasmid- and chromosome-mediated beta- Faecalibacterium as well as the search for genes codifying for the
lactamases (Wiseman et al., 1997), with a MIC higher than most important resistance mechanisms for, at least, some of the
256 mg/L. antibiotics tested in this study.
The analysis of antimicrobial resistance is of major
importance due to the fast evolution of antibiotic resistance in Metabolic Activities
response to the extensive use of antimicrobials. However, the Enzymatic activities detected by API-20A gallery system are
microbiological breakpoints marked by the EFSA for most of reported in Table 4. Interestingly, only one enzyme was detected
Gram positive bacteria is probably not the most correct for the and active in all the tested strains: the beta-galactosidase.
analysis of F. prausnitzii isolates as no so many information Otherwise, all the strains were not able to ferment mannose
about their natural or acquired resistance patters is reported, to or raffinose, to reduce nitrate and to produce indole (data not
our knowledge, up to day in the literature. Foditsch et al. (2014) shown). Furthermore, all the isolates were negative for the
have identify that more of the 50% of the F. prausnitzii strains presence of urease, arginine dihydrolase, beta-glucosidase, alpha-
that they isolated from fecal samples of healthy calves and piglets arabinosidase, N-acetyl-beta-glucosaminidase, glutamic acid
were resistant to tetracycline, amikacin, cefepime, and cefoxitin decarboxylase, alkaline phosphatase, phenylalanine arylamidase,
comparing the MIC values with the standard values determined leucine arylamidase, pyroglutamic acid arylamidase, tyrosin
by CLSI for Bacteroides fragilis ATCC 25285. This fact highlights arylamidase, alanine arylamidase, glutamyl glutamic acid
the need of more microbiological studies of antibiotic resistance arylamidase, and serin arylamidase (data not shown). These
in this species in order to determine a correct standard values for results confirm previous observations where no strain was able
the same metabolic profile and donor. And the third metabolic
II), which are the only ones resistant to cefpirome, share also
strains CNCM I-4543 and CNCM I-4574 (group B, phylogroup
the group A from phylogroup I (CNCM I-4546 and M21/2). The
included in three different profiles. Two of them corresponds to
While six strains showed individual profiles, the other seven are
reported in some F. prausnitzii isolates (Lopez-Siles et al., 2012).
histidine-arylamidase), differences inter-strains were detected
leucyl glycerine-arylamidase, glycine-arylamidaseycine, and
alpha-glucosidase, beta-glucuronidase, arginine arylamiase,
For all the other enzymes (6 phospho-beta galactosidase,
8
TABLE 3 | Minimum inhibitory concentrations (MIC) (mg/L) for the different antibiotics tested.
EFSA Breakpoint GEN STR KM ERY CLI VAN TET QD CM AMP CPO LZD
(other Gram +)
A2-165 1.37 ± 0.12 96 ± 18.47 8.67 ± 1.76 0.20 ± 0.08 0.016 ± 0 0.27 ± 0.05 0.016 ± 0 0.03 ± 0.01 0.08 ± 0.02 0.11 ± 0.02 9±1 1.25 ± 0.25
L2-6 4.33 ± 0.88 32 ± 7.15 0.42 ± 0.08 0.10 ± 0.02 0.023 ± 0 0.47 ± 0.12 4.21 ± 0.62 0.58 ± 0.46 20 ± 5.66 0.06 ± 0.03 24 ± 4.62 0.62 ± 0.12
M21/2 4.75 ± 1.49 21 ± 4.43 51 ± 45 0.05 ± 0.01 0.016 ± 0 0.25 ± 0 1.09 ± 0.63 0.016 ± 0 0.15 ± 0.03 0.04 ± 0.03 16.67 ± 7.86 0.75 ± 0
CNCM I-4540 1.25 ± 0.25 21.33 ± 2.67 122.67 ± 66.83 0.11 ± 0.02 0.016 ± 0 0.71 ± 0.18 0.016 ± 0 0.011 ± 0.004 0.05 ± 0.04 0.11 ± 0.02 24 ± 4.62 0.04 ± 0.06
CNCM I-4541 1.62 ± 0.47 24 ± 0 14 ± 2 20 ± 4 0.016 ± 0 0.125 ± 0 8±0 0.27 ± 0.24 0.17 ± 0.07 0.06 ± 0 40 ± 8 1.25 ± 0.25
CNCM I-4542 2.87 ± 1.12 23.67 ± 3.26 20 ± 4 0.11 ± 0.07 0.016 ± 0 0.67 ± 0.17 0.016 ± 0 0.016 ± 0 0.29 ± 0.40 0.08 ± 0.03 20.67 ± 7.69 3.17 ± 1.42
CNCM I-4543 2 ± 0.40 16 ± 0 20 ± 4 0.08 ± 0.02 0.016 ± 0 0.29 ± 0.04 0.2 ± 0.002 0.07 ± 0.03 0.31 ± 0.09 0.12 ± 0 ≥256 ± 0 1.25 ± 0.38
CNCM I-4544 6 ± 1.15 21.2 ± 7.31 100 ± 52.41 0.12 ± 0 0.016 ± 0 0.92 ± 0.08 0.016 ± 0 0.023 ± 0 0.11 ± 0.02 0.09 ± 0.02 53.33 ± 5.33 0.5 ± 0
CNCM I-4546 10 ± 66 50.67 ± 13.33 256 ± 0 0.22 ± 0.61 0.018 ± 0.002 0.56 ± 0.31 2.6 ± 0.6 0.16+0.07 1.25 ± 1.51 0.07 ± 0.02 24.8 ± 6.24 3±1
CNCM I-4573 7±1 32 ± 0 234 ± 21.33 0.01 ± 0.03 0.024 ± 0.007 0.28 ± 0.05 0.028 ± 0.003 0.04 ± 0.005 0.38 ± 0 0.25 ± 0 22 ± 3.83 3.3 ± 0.8
CNCM I-4574 1.75 ± 0.25 14 ± 0 6±2 0.07 ± 0.02 0.016 ± 0 0.25 ± 0 0.016 ± 0 0.03 ± 0.01 0.09 ± 0.02 0.22 ± 0.03 ≥256 ± 0 0.5 ± 0.14
CNCM I-4575 1.25 ± 0.25 5±1 4±1 0.07 ± 0.01 0.016 ± 0 0.5 ± 0 0.032 ± 0 0.03 ± 0.03 0.016 ± 0 0.084 ± 0.02 4.67 ± 1.67 0.03 ± 0.01
CNCM I-4644 0.91 ± 0.08 9.33 ± 1.33 135 ± 69.84 0.10 ± 0.05 0.026 ± 0.01 0.23 ± 0.02 0.83 ± 0.32 0.04 ± 0.01 0.079 ± 0.04 0.026 ± 0.01 9.333 ± 1.33 0.58 ± 0.08
Gentamicin (GEN), streptomycin (STR), kanamycin (KM), erythromycin (ERY), clindamycin (CLI), vancomycin (VAN), tetracycline (TET), quinupristin/dalfopristin (QD), chloramphenicol (CM), ampicillin (AMP), cefpirome (CPO), and linezolid
(LZD). Experiments have been done in triplicate and the results are expressed as the media ± SEM. nd, Not defined. In bold, Resistances.
Functional Characterization of F. prausnitzii Strains
beta− beta Galactosidase Alpha beta Arginine Leucyl Glycine Glycine Histidine
galactosidase 6 phosphate glucosidase glucuronidase Arylamidase Arylamidase Arylamidaseycine Arylamidase
A2−165 + + + + + + + +
L2−6 + − + + + + + +
M21/2 + − − − + − + +
CNCM I−4540 + + − − − − − −
CNCM I−4541 + + − − + − − −
CNCM I−4542 + + − − − − − −
CNCM I−4543 + + − + + + + −
CNCM I−4544 + − − − − − − −
CNCM I−4546 + − − − + − + +
CNCM I−4573 + − − − + − + +
CNCM I−4574 + + − + + + + −
CNCM I−4575 + + − + + − + +
CNCM I−4644 + − − − + + + +
profile is shared by strains CNCM I-4540 and CNCM I-4542 that major component of the microbiota, could not have metabolic
belong to the group C of phylogroup II. deleterious impact on the host.
It is now well-establish that F. prausnitzii is an acetate- All strains were unable to growth in the presence of mucin
consumer and butyrate-producer species (Duncan et al., 2002; as the only carbon source in a defined medium (data not
Lopez-Siles et al., 2012). Here, we report that in pure cultures, shown). This data agrees with previous results where no evidence
our new isolated strains are also able to produce butyrate and of fermentation of porcine gastric mucin by F. prausnitzii
this production is significantly and positively correlated to their was detected (Lopez-Siles et al., 2012). Nevertheless, SCFA
growth (OD600nm ; r = 0.8462; p = 0.003; Figures 5A,B). It is concentrations and OD600nm measures taken after 2 days of
interesting to highlight that the production level of butyrate was incubation showed the ability of the different strains to survive
not linked to a particular phylogroup (Phylogroup I 3.91 mM ± but metabolically inactive as it could be deduced by the absence
0.43 and Phylogroup II 4.89 mM ± 0.62). Moreover, all strains of butyrate in the supernatants of the cultures and the almost
could metabolize acetate present in the culture medium at around minimal OD600nm recorded (data no shown). A decrease in
the same level (Figure 5A). This consumption was not directly butyrate production due to non-optimal growth conditions have
correlated to bacterial growth (r = −0.3132, p = 0.2975) and been already reported for F. prausnitzii A2-165 strain (Lopez-
tended to be more correlated to butyrate production (r = −0.544, Siles et al., 2012). This characteristic pointed out the intrinsic
p = 0.0546). This observation is in agreement with the literature growth requirements of this species which, in addition to be an
which describes that most of the carbon present in the butyrate EOS, needs strain specific nutritional environment and has the
produced (around 85%) is derived from external acetate, with ability to switch between substrates derived from the diet or the
only 15% provided directly from glucose (FEEDAP, 2012). host (Lopez-Siles et al., 2017).
F. prausnitzii can also produce a few amount of D-lactate
(FEEDAP, 2012). Indeed, among our strain collection, no L- Lytic Activities
lactate was detected and only few amounts of D-lactate were Gelatin is a heterogeneous mixture of water-soluble protein
detected (1.09 mM ± 0.15 and 1.07 mM ± 0.39 phylogroup I and that is usually used in microbiological procedures to detect the
II respectively; data not shown). This production, not correlated presence of proteolytic activities. None of the strains were able
with phylogroup affiliation, was correlated to the OD600nm (r to degrade gelatin in the conditions recommended by the API
= 0.6209, p = 0.0235). Bacterial D-Lactate production can be gallery supplier (data not shown). However, when the strains
viewed as harmful since accumulation of this metabolite into the were inoculated in the gallery in a defined medium instead of
blood may be neurotoxic and leads to acidosis (Mack, 2004). API suspension medium, they were able to degrade partially this
In particular, humans with short bowel syndrome (in which compound after 3 days of incubation. This fact suggests that the
small intestine has been surgically removed), the D/L fecal lactate strains are able to hydrolyze gelatin although, maybe due to the
ratio seems to be the most relevant index with a higher D- growth limitations present in this culture media, the existence of
encephalopathy risk (Mayeur et al., 2013). However, in healthy this compound is not enough to allow the metabolic development
adults, there is no lactate detectable in fecal samples, because of this activity in the strains.
lactobacilli (main producer of D-lactate) are minor groups in The presence of hemolytic activity was tested using blood
microbiota and lactate is degraded by other major bacterial agar plates. None of the strains showed hemolytic activity
groups (36, He, 2008 #41). This observation also suggested under the conditions tested. In contrast, all the strains reveal
that the weak production of D-lactate by F. prausnitzii strains, a DNAse activity in green methyl-DNA medium (data not
FIGURE 5 | SCFA metabolism of F. prausnitzii strains in vitro. (A) Acetate and butyrate concentrations after 20 h growth of strain in YBHI supplemented medium. The
concentration of control media was subtracted for each measurement have been done at least in triplicate. (B) Correlation between OD600 nm and butyrate production.
shown). Furthermore, the presence of a magnesium dependent For the PBMC assay, although all the strains tend to increase
DNase activity has been previously reported in at least three the production of IL-10 cytokine, only four strains (two controls
of five strains already sequenced [A2-165 (gi:257439194), SL3/3 and two new isolates from this study) were able to induce
(gi:295105207), and L2/6 (gi:295102777)]. statistically significant increase production of this cytokine
The presence of these extracellular activities is often linked to (Figure 6B) The two most performing strains (A2-165 and
a virulence status in some bacterial species such as Enterococcus 4543) belong to the phylogroup II, group B. Notably, the IL-
spp. (Eaton and Gasson, 2001). However, these factors also 10 production was correlated with both growth ratio (r =
contribute to the survival of microorganisms in the mammalian 0.6813, p = 0.0103) and butyrate production (r = −0.6923, p
gut being characteristic of several members of the natural = 0.0126). This different phenotype may suggest the presence of
microbiota (Sanders et al., 2010). This can be the case of different molecule(s) responsible of the anti-inflammatory effects
Faecalibacterium isolates, which are extremely well-adapted to in vitro. The anti-inflammatory properties of butyrate have been
the gut environment (Lopez-Siles et al., 2012). already reported in the literature (Fusunyan et al., 1999; Kamitani
et al., 1999) and its ability to block IL-8 production under the
Antibacterial Activities conditions tested in this study were confirmed in vitro in similar
We investigated antibacterial properties of F. prausnitzii concentrations to those founds in F. prausnitzii supernatants
supernatants, using the bacteriocin activity assay. We did not (data not shown). However, its role remains controversial as
reveal any antibacterial effect on several anaerobic and aerobic its effects seems to be dose- and time-dependent as well as
bacterial species under the conditions tested. This fact is a depended on the cellular model used (Martin et al., 2013).
desirable characteristic of a strain to be considered as a probiotic For instance, regarding cells from intestinal origin, butyrate
candidate. has been found to decrease the secretion of IL-8 in Caco-2
and HIPEC cells and, in contrast to this study, to enhance IL-
Ability to Stimulate the Immune Response 8 production in HT-29 and HT-29 MTX cells (Bocker et al.,
The reference strain F. prausnitzii A2-165 is well-known for 2003).
its immuno-modulatory properties and more specifically for its However, several authors have found different candidate
anti-inflammatory effects both in vitro and in vivo in different molecules/structures responsible for F. prausnitzii anti-
murine models of colitis (Sokol et al., 2008; Martin et al., 2014). inflammatory effects. MAM protein, found in F. prausnitzii
To determine whether the newly isolated F. prausnitzii strains supernatant, has been found to block NF-κB activation and the
are able to modulate the immune response, we tested in vitro production of the pro-inflammatory cytokine IL-8 (Quevrain
the immuno-modulatory properties of the supernatants from all et al., 2016). F. prausnitzii is also able to produce bioactive
the isolates in two different cellular models: HT-29 and PBMC. anti-inflammatory molecules such as shikimic and salicylic acids
The first one is based on the capacity to block IL-8 production (Miquel et al., 2015b). Besides, Rossi and co-workers showed
(a pro-inflammatory cytokine) induced by TNF-α stimulation in the ability of F. prausnitzii strain HTF-F and its extracellular
HT-29 epithelial cells and the second is based on the stimulation polymeric matrix to develop immunomodulatory effects through
of PBMC cells and the measure of the anti-inflammatory cytokine the TLR2 dependent modulation of IL-12 and IL-10 cytokine
IL-10. As shown in Figure 6A, all the strains tend to decrease IL-8 production in human monocyte-derived dendritic cells (Rossi
concentrations. However, this decrease was not equivalent in all et al., 2015) and F. prausnitzii has been found to be a strong
the strains and does not correlate either with growth ratio (r = inducer of regulatory T cells secreting IL-10 (Sarrabayrouse et al.,
−0.2857, p = 0.344) or butyrate production (r = −0.3357, p = 2014). All these results point out the complex anti-inflammatory
0.2869). mechanisms underlying this species.
CONCLUDING REMARKS
The development of new probiotic products containing human
isolated strains with beneficial properties for the host requires
the development of new techniques in order to: (i) isolate strains
belonging to the major groups of the intestinal microbiota, (ii)
determinate their safe status and (iii) infer in their potential
FIGURE 6 | Immuno-modulation capacities of F. prausnitzii strains in vitro.
(A) IL-8 production in HT-29 TNF-α stimulated cells. Experiments have been
beneficial effects. This study meets these entire three requests.
done at least in triplicate. Results are expressed as IL-8/ protein (pg/mg) and Work with anaerobic and more precisely EOS bacteria are a
have been normalized using as reference value the IL-8 produced after the prerequisite to succeed in the isolation of representative strains
co-incubation with PBS as a negative control. (B) IL-10 production in that can impact on intestinal homeostasis. For this reason, in this
peripheral blood mononuclear cells. Experiments have been done at least in
study, we have used a new procedure to isolate EOS strains from
triplicate. Results are expressed as IL-10 concentration (pg/mL). Significant
differences from the control (YBHI) was specified as: *p < 0.05 and ***p < feces that has enabled us to build a collection of F. prausnitzii
0.001. strains. The lack of knowledge about this species prompts us
to further analyze their genetic diversity by comparing the new
isolates with those already available in the databases. This has
allowed us to point out the high diversity of our collection
Adhesion to Epithelial Cells In vitro ranged on two different phylogroups with different clusters.
In parallel, we also sought for the adhesion capacities of the F. prausnitzii strain genomes should be established or/and a
new F. prausnitzii isolates to the intestinal epithelial cells HT- metabolic comparison of several strains in the same culture
29 and mucin. All the tested strains were not able to adhere to conditions whether the phylogroups belong to genomovars or
HT-29 cells in vitro (data not shown) in anaerobic conditions. genomospecies.
Regarding mucin, some of the strains were able to adhere to Regarding safety concerns, this study is the first step toward
this compound after 3 h of incubation in the anaerobic chamber a better understanding of F. prausnitzii properties. Up to date,
(Figure 7), Even if our conditions were not representative of little was known about F. prausnitzii resistance to antibiotics, lytic
physiological conditions (death of our eukaryotic cells), this activities or adhesion properties. Here, we have shown for the
result gives ecological clues about the processes of colonization first time the profile of all these characteristics in a collection of
of the gastro-intestinal tract by F. prausnitzii. In fact, this species human Faecalibacterium strains. A positive remark is that all the
is a late but major commensal colonizer of the gut which strains were not antibacterial producers, not hemolytic and weak
producer of D-lactate. Furthermore, although some of the strains performed the experiments and analysis. RM and SM draft the
were able to adhere to mucin, this trait can be considered as factor manuscript. VR, CB, FC, JC, HS, LGBH, MT, and PL revised the
favoring durable implantation and a highly effective probiotic manuscript critically. All the authors have read and approved the
(Miquel et al., 2015a). However, further analyses are required to last version of the manuscript.
better determine the presence of acquired or natural resistances
as well as to distinguish between the pathogenic or adaptative FUNDING
nature of some of the properties detected such as the presence
of DNase activity. This paper was a part of FPARIS collaborative project selected
Finally, the anti-inflammatory properties of all the strains and supported by the Vitagora Competitive Cluster and
have been analyzed. There is a well-known correlation between funded by the French FUI (Fond Unique Interministériel; FUI:
F. prausnitzii dysbiosis and a large set of human diseases such n◦ F1010012D), the FEDER (Fonds Européen de Développement
as IBD and IBS (Miquel et al., 2013). Recent studies using Régional; Bourgogne: 34606), the Burgundy Region, the Conseil
F. prausnitzii strains in in vivo models provide arguments Général 21 and the Grand Dijon. This work was also supported
concerning its beneficial effect on the host (Sokol et al., 2008; by Merck Médication Familiale (Dijon, France) and Biovitis
Wrzosek et al., 2013; Martin et al., 2014). The presence of the anti- (Saint Etienne de Chomeil, France). RM and SM receive a salary
inflammatory properties of these strains also opens the possibility from the same grants.
to test them in murine models in order to further determine
their beneficial effects before testing them in human clinical
ACKNOWLEDGMENTS
trials.
Authors thank Prof. Juan Evaristo Suárez for the critical reading
AUTHOR CONTRIBUTIONS of the manuscript and Stéphanie Courau and Pascal Molimard
for fruitful discussions during the project. We gratefully
RM, SM, JC, HS, LGBH, MT, and PL participate in the design acknowledge T. Meylheuc for scanning electron microscopy
of the project. RM, SM, JC, HS, OB, VA, LGBH, MT, and PL (MIMA2 platform, INRA, France) and Harry Flint for the
designed the experiments. RM, SM, LB, CB, VR, SH, and FC reference Faecalibacterium strains.
Kimura, M. (1980). A simple method for estimating evolutionary rates of base specific molecular determinants. Appl. Microbiol. Biotechnol. 100, 9605–9617.
substitutions through comparative studies of nucleotide sequences. J. Mol. Evol. doi: 10.1007/s00253-016-7813-0
16, 111–120. doi: 10.1007/BF01731581 Radziwill-Bienkowska, J. M., Zochowska, D., Bardowski, J., Mercier-Bonin, M.,
Lan, A., Bruneau, A., Bensaada, M., Philippe, C., Bellaud, P., Rabot, S., and Kowalczyk, M. (2014). Lactococcus lactis IBB477 presenting adhesive and
et al. (2008). Increased induction of apoptosis by Propionibacterium muco-adhesive properties as a candidate carrier strain for oral vaccination
freudenreichii TL133 in colonic mucosal crypts of human microbiota- against influenza virus. Acta Biochim. Pol. 61, 603–607.
associated rats treated with 1,2-dimethylhydrazine. Br. J. Nutr. 100, 1251–1259. Rajilic-Stojanovic, M., Biagi, E., Heilig, H. G., Kajander, K., Kekkonen, R. A.,
doi: 10.1017/S0007114508978284 Tims, S., and de Vos, W. M. (2011). Global and deep molecular analysis
Lopez-Siles, M., Duncan, S. H., Garcia-Gil, L. J., and Martinez-Medina, M. of microbiota signatures in fecal samples from patients with irritable bowel
(2017). Faecalibacterium prausnitzii: from microbiology to diagnostics and syndrome. Gastroenterology 141, 1792–1801. doi: 10.1053/j.gastro.2011.07.043
prognostics. ISME J. 11, 841–852. doi: 10.1038/ismej.2016.176 Ramirez-Farias, C., Slezak, K., Fuller, Z., Duncan, A., Holtrop, G., and Louis,
Lopez-Siles, M., Khan, T. M., Duncan, S. H., Harmsen, H. J., Garcia-Gil, L. J., P. (2009). Effect of inulin on the human gut microbiota: stimulation of
and Flint, H. J. (2012). Cultured representatives of two major phylogroups of Bifidobacterium adolescentis and Faecalibacterium prausnitzii. Br. J. Nutr. 101,
human colonic Faecalibacterium prausnitzii can utilize pectin, uronic acids, 541–550. doi: 10.1017/S0007114508019880
and host-derived substrates for growth. Appl. Environ. Microbiol. 78, 420–428. Rossi, O., Khan, M. T., Schwarzer, M., Hudcovic, T., Srutkova, D., Duncan, S. H.,
doi: 10.1128/AEM.06858-11 et al. (2015). Faecalibacterium prausnitzii strain HTF-F and its extracellular
Louis, P., Duncan, S. H., McCrae, S. I., Millar, J., Jackson, M. S., and Flint, H. J. polymeric matrix attenuate clinical parameters in DSS-induced colitis. PLoS
(2004). Restricted distribution of the butyrate kinase pathway among butyrate- ONE 10:e0123013. doi: 10.1371/journal.pone.0123013
producing bacteria from the human colon. J. Bacteriol. 186, 2099–2106. Sanders, M. E., Akkermans, L. M., Haller, D., Hammerman, C., Heimbach, J.,
doi: 10.1128/JB.186.7.2099-2106.2004 Hormannsperger, G., et al. (2010). Safety assessment of probiotics for human
Mack, D. R. (2004). D(-)-lactic acid-producing probiotics, D(-)-lactic acidosis and use. Gut Microbes 1, 164–185. doi: 10.4161/gmic.1.3.12127
infants. Can. J. Gastroenterol. 18, 671–675. doi: 10.1155/2004/342583 Sarrabayrouse, G., Bossard, C., Chauvin, J. M., Jarry, A., Meurette, G., Quevrain, E.,
Martin, R., Chain, F., Miquel, S., Lu, J., Gratadoux, J. J., Sokol, H., et al. (2014). et al. (2014). CD4CD8αα lymphocytes, a novel human regulatory T cell subset
The commensal bacterium Faecalibacterium prausnitzii is protective in DNBS- induced by colonic bacteria and deficient in patients with inflammatory bowel
induced chronic moderate and severe colitis models. Inflamm. Bowel Dis. 20, disease. PLoS Biol. 12:e1001833. doi: 10.1371/journal.pbio.1001833
417–430. doi: 10.1097/01.MIB.0000440815.76627.64 Schmidt, C. (2013). The startup bugs. Nat. Biotechnol. 31, 279–281.
Martin, R., Miquel, S., Ulmer, J., Kechaou, N., Langella, P., and Bermudez- doi: 10.1038/nbt.2544
Humaran, L. G. (2013). Role of commensal and probiotic bacteria in human Sokol, H., Pigneur, B., Watterlot, L., Lakhdari, O., Bermudez-Humaran, L.
health: a focus on inflammatory bowel disease. Microb. Cell Fact. 12:71. G., Gratadoux, J. J., et al. (2008). Faecalibacterium prausnitzii is an anti-
doi: 10.1186/1475-2859-12-71 inflammatory commensal bacterium identified by gut microbiota analysis
Martin, R., Sanchez, B., Urdaci, M. C., Langella, P., Suarez, J. E., and Bermudez- of Crohn disease patients. Proc. Natl. Acad. Sci. U.S.A. 105, 16731–16736.
Humaran, L. G. (2015). Effect of iron on the probiotic properties of the vaginal doi: 10.1073/pnas.0804812105
isolate Lactobacillus jensenii CECT 4306. Microbiology 161(Pt 4), 708–718. Tamura, K., Stecher, G., Peterson, D., Filipski, A., and Kumar, S. (2013). MEGA6:
doi: 10.1099/mic.0.000044 molecular evolutionary genetics analysis version 6.0. Mol. Biol. Evol. 30,
Mayeur, C., Gratadoux, J. J., Bridonneau, C., Chegdani, F., Larroque, B., Kapel, 2725–2729. doi: 10.1093/molbev/mst197
N., et al. (2013). Faecal D/L lactate ratio is a metabolic signature of microbiota Thompson, J. D., Higgins, D. G., and Gibson, T. J. (1994). CLUSTAL W: improving
imbalance in patients with short bowel syndrome. PLoS ONE 8:e54335. the sensitivity of progressive multiple sequence alignment through sequence
doi: 10.1371/journal.pone.0054335 weighting, position-specific gap penalties and weight matrix choice. Nucleic
Miquel, S. R., Martìn, Bridonneau, C., Robert, V., Sokol, H., Bermúdez-Humarán, Acids Res. 22, 4673–4680. doi: 10.1093/nar/22.22.4673
L., Thomas, M., et al. (2014). Ecology and metabolism of the beneficial intestinal van Tongeren, S. P., Slaets, J. P., Harmsen, H. J., and Welling, G. W. (2005). Fecal
commensal bacterium Faecalibacterium prausnitzii. Gut Microbes 5, 146–151. microbiota composition and frailty. Appl. Environ. Microbiol. 71, 6438–6442.
doi: 10.4161/gmic.27651 doi: 10.1128/AEM.71.10.6438-6442.2005
Miquel, S., Beaumont, M., Martin, R., Langella, P., Braesco, V., and Thomas, Wiseman, L. R., Lamb, H., and Cefpirome, M. (1997). A review of its antibacterial
M. (2015a). A proposed framework for an appropriate evaluation scheme for activity, pharmacokinetic properties and clinical efficacy in the treatment of
microorganisms as novel foods with a health claim in Europe. Microb. Cell Fact. severe nosocomial infections and febrile neutropenia. Drugs 54, 117–140.
14, 48. doi: 10.1186/s12934-015-0229-1 doi: 10.2165/00003495-199754010-00013
Miquel, S., Leclerc, M., Martin, R., Chain, F., Lenoir, M., Raguideau, S., Wrzosek, L., Miquel, S., Noordine, M. L., Bouet, S., Chevalier-Curt, M. J.,
et al. (2015b). Identification of metabolic signatures linked to anti- Robert, V., et al. (2013). Bacteroides thetaiotaomicron and Faecalibacterium
inflammatory effects of Faecalibacterium prausnitzii. Mbio 6:e00300-15. prausnitzii influence the production of mucus glycans and the development of
doi: 10.1128/mBio.00300-15 goblet cells in the colonic epithelium of a gnotobiotic model rodent. BMC Biol.
Miquel, S., Martin, R., Rossi, O., Bermudez-Humaran, L., Chatel, J., Sokol, H., et al. 11:61. doi: 10.1186/1741-7007-11-61
(2013). Faecalibacterium prausnitzii and human intestinal health. Curr. Opin.
Microbiol. 16, 255–261. doi: 10.1016/j.mib.2013.06.003 Conflict of Interest Statement: PL and HS are co-founders of the start-up
Neish, A. S. (2009). Microbes in gastrointestinal health and disease. NextBiotiX aiming to use next-generation probiotics to fight and to prevent IBD.
Gastroenterology 136, 65–80. doi: 10.1053/j.gastro.2008.10.080
Pineiro, M., and Stanton, C. (2007). Probiotic bacteria: legislative framework– The other authors declare that the research was conducted in the absence of
requirements to evidence basis. J. Nutr. 137(3 Suppl. 2), 850S–853S. any commercial or financial relationships that could be construed as a potential
Qin, J., Li, R., Raes, J., Arumugam, M., Burgdorf, K. S., Manichanh, C., et al. conflict of interest.
(2010). A human gut microbial gene catalogue established by metagenomic
sequencing. Nature, 464, 59–65. doi: 10.1038/nature08821 Copyright © 2017 Martín, Miquel, Benevides, Bridonneau, Robert, Hudault, Chain,
Quevrain, E., Maubert, M. A., Michon, C., Chain, F., Marquant, R., Tailhades, J., Berteau, Azevedo, Chatel, Sokol, Bermúdez-Humarán, Thomas and Langella. This
et al. (2016). Identification of an anti-inflammatory protein from is an open-access article distributed under the terms of the Creative Commons
Faecalibacterium prausnitzii, a commensal bacterium deficient in Crohn’s Attribution License (CC BY). The use, distribution or reproduction in other forums
disease. Gut 65, 415–425. doi: 10.1136/gutjnl-2014-307649 is permitted, provided the original author(s) or licensor are credited and that the
Radziwill-Bienkowska, J. M., Le, D. T., Szczesny, P., Duviau, M. P., Aleksandrzak- original publication in this journal is cited, in accordance with accepted academic
Piekarczyk, T., Loubiere, P., et al. (2016). Adhesion of the genome- practice. No use, distribution or reproduction is permitted which does not comply
sequenced Lactococcus lactis subsp. cremoris IBB477 strain is mediated by with these terms.